Image Analysis for Diagnostics of Insulators with Biological Contamination

Licentiate thesis, 2002

Composite insulators have a number of advantages compared to conventional insulators. However, the fact that they are more prone to aging is a major drawback. For instance, biological growths, like algae and fungi, have been found to colonize composite insulators, and possibly effecting the performance of the materials. So far there is a lack of diagnostic methods capable of detecting material changes of composite insulators in the field, and developing new diagnostic methods has thus become very important. One of the main aims of the work presented in this thesis was to elucidate on the influence of biological contamination on the electrical performance of composite insulators. To do this, fungi were grown on silicone rubber insulators and materials samples in a climate chamber, and leakage current measurements were performed occasionally during a period of one year. The measurements were performed under clean fog conditions when applying ac-voltage. In parallel photos of the surfaces of samples and insulators covered with the growth were taken. To characterize the distribution of this contamination, tools based on digital image analysis were applied. The first experiments were performed on images of flat material samples with fungal growth, but later the investigations were expanded to parts of real insulators as well. From the current measurements, it was found that the contaminated insulators always suffered from higher leakage currents than the clean ones. Moreover, discharge activity could be observed on the insulators with fungal growth. Regarding the application of digital image analysis, parameters like total covered area, area of largest cluster and number of clusters were found to be useful in characterizing the distributions. Finally, methods for estimating the length of creepage distance not covered by growth were developed.